Cent. Eur. J. Phys. • 9(2) • 2011 • 344-348 DOI: 10.2478/s11534-010-0106-9

Central European Journal of Physics

Secondary electron microscopy and transmission electron microscopy studies of carbon nanotubes in C-Ni films Research Article

Mirosław Kozłowski1∗ , Piotr Dłużewski2 , Ewa Kowalska1 , Elżbieta Czerwosz1 1 Tele & Radio Research Institute, ul. Ratuszowa 11, 03-450 Warszawa, Poland 2 Institute of Physics PAS, Al. Lotników 32/46, 02-668 Warszawa, Poland

Received 30 July 2010; accepted 12 October 2010

Abstract:

Carbon nanotubes films have been studied with SEM and TEM. The studied films were obtained using a two step method: PVD process and CVD process. Strongly defected and curled carbon nanotubes containing Ni nanoparticles formed the film with thickness of about 300-400 nm. Observed carbon nanotubes were of lengths from 100 nm to 300 nm and did not stick to each other.

PACS (2008): 68.37.-d Keywords:

MWCNT • PVD • CVD • nickel nanoparticles © Versita Sp. z o.o.

1.

Introduction

Multiwall carbon nanotubes (MWCNT) films are obtained by many methods with various catalysts, but one of the most frequently used is nickel [1–9]. The most popular method is catalytic chemical vapor deposition (CCVD) [7]. The structure, length and type of carbon nanotubes (CNT) depend on the kind and parameters of this technological process. When the catalyst remains anchored to the substrate, the synthesis mode is called “base-growth”. On the other hand, the growth follows a “tip-growth” mechanism when the particle lifts off the substrate and is observed at the top of the CNTs [2–4]. Many authors discuss different ∗

344

E-mail: [email protected]

reasons for these growth mechanisms. The Tele & Radio Research Institute developed an original two-step method for synthesis of carbon nanotubes films in which single nanotubes were placed at large separations from each other on the film surface (rarely dispersed placement of CNTs) [10]. This kind of nanotube placement could prevent the existence of an electrostatic repulsion effect between nanotubes. Such films could find an application as effective field emitters.

2.

Experimental

MWCNTs films were obtained by a two step method. The first step was PVD (Physical Vapor Deposition) process, and the second CVD (Chemical Vapor Deposition) pro-

Mirosław Kozłowski, Piotr Dłużewski, Ewa Kowalska, Elżbieta Czerwosz

cess. In the first step, a carbonaceous matrix film in which nickel nanoparticles were placed was deposited on the Si substrate. The PVD process was performed from two separated sources containing fullerene C60 and nickel acetate under a dynamic vacuum of 10−5 mbar. As a result a polycrystalline C-Ni film was obtained (Fig. 1). The concentration of Ni in the film was 69% by mass. Figure 2.

Figure 1.

SEM image of polycrystalline C-Ni film from PVD. In the inset, a TEM image and selected area diffraction pattern showing diffuse rings from Ni fcc nanocrystals structure can be seen.

The granular polycrystalline C-Ni film from the PVD process was used as the catalyst support in the second step (CVD) in which the pyrolysis of xylene (C8 H10 ) proceeded. The scheme of experimental CVD set-up can be seen in Fig. 2. The process parameters were as follows: • temperature of pyrolysis process - 650°C, • argon flow - 40 l/h, • xylene flow - 0.1 ml/min, • duration - 10 min. As a result of the CVD process a new film consisting of many carbon forms, Ni nanoparticles and the multi-walled carbon nanotubes (MWCNTs) was formed. The CVD process parameters were chosen in such way as to obtain a film with carbon nanotubes rarely distributed on its surface. Such films, in which carbon nanotubes are placed with a large distance between each other are good candidates for effective field emitters. In the case where MWCNT are highly packed on a film surface, electrostatic shielding may cause a reduction of field emission enhancement factor. Therefore, to obtain rarely displaced MWCNT in the film, the form of initial PVD film with small

CVD set-up.

diameter Ni nanocrystals (a few nm) was used, while the CVD process was carried out using a low dose of xylene, and over the short time of 10 minutes. The morphology and structure of these films was studied by Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) methods. SEM investigation was performed with the JEOL JSM-7600F field emission scanning electron microscope, operating at 5 keV incident energy. SEM studies were carried out using two modes of operation: SE (Secondary Electron) and LABE (Low Angle Backscattered Electron). In this mode, the contrast depends on the atomic number of elements of the observed microstructure. Areas with higher atomic number are mapped as bright, with those of lower atomic number shown as darker. TEM was applied to study the structure of the deposited film. TEM investigations were performed with the JEOL JEM-2000EX electron transmission microscope operating at 200 keV incident electron beam energy.

3.

Results

The aim of these studies was to determine an areal density of MWCNTs, their structure and sizes. From SEM images, the length and diameter of MWCNTs, the size of Ni nanoparticles and the distribution of MWCNTs on the film surface were determined. From TEM images and selected area electron diffraction (SAED), MWCNTs and Ni nanoparticle structure was determined.

4.

SEM investigations

SEM studies showed that film adhesion to the substrate is low (Fig. 3a). The shape of film surface (highly folded) suggests that the stress generated during the CVD process could be responsible for the formation of such surface. However, no cracks in the surface are observed. Short time of duration with a strong flow of argon and xylene in the CVD process caused the growth of carbon nanotubes

345

Secondary electron microscopy and transmission electron microscopy studies of carbon nanotubes in C-Ni films

with low areal density (Fig. 3b). Observed MWCNTs are short (100 ÷ 300 nm) and do not stick to each other.

Figure 4.

Histogram of the MWCNTs size.

is ∼ 30 − 40 nm while for the elongated Ni nanoparticles placed within MWCNTs the length is ∼ 40 − 50 nm and the diameter of such a cylinder is ∼ 15 nm.

5.

Figure 3.

Investigated layer in different magnifications a) 100×, b) 10000×.

The MWCNTs diameters are presented in Fig. 4. The diameters of the carbon nanotubes are between 30-70 nm, with the average diameter being 49.7 nm (Fig. 4). SEM investigations of the film at high magnifications show the position of Ni nanoparticles in MWCNTs (Fig. 5a). In LABE mode (Fig. 5b) Ni nanoparticles are clearly visible as bright objects. These objects are placed on top of carbon nanotubes and inside the carbon nanotubes. In contrary to the objects place on the top of MWCNT, the shape of these last objects is elongated. Comparison of SE and LABE images shows differences in diameters of the Ni nanoparticles detected by secondary and backscattered electrons detectors. It suggests that pure Ni nanoparticles are covered with carbon. From LABE images we can determine the size of pure Ni nanoparticles. We found that for Ni nanoparticles from the top of MWCNTs the diameter

346

TEM investigations

TEM investigation allowed the structure of Ni nanoparticles enclosed inside MWCNTs and on their tops (Fig. 6) to be determined. Merkulov et al. [3] describe a mechanism of tip and base growth. In the first case Ni nanoparticles are found at the base of nanotubes and in the second case they are located on a top of nanotubes. They suggest that the placement of Ni nanoparticles is connected to electrostatic force, forming a uniform tensile stress across an entire particle/CNT interface. When the particle is at the top, the electrostatic force produces a compressive force at the CNT/particle interface where a greater growth rate is observed; on the side where less growth rate happens, a tensile stress is applied at the interface. This opposite behavior favors subsequent carbon precipitation at the interface with tensile stress (and the smaller rate of growth). The net result is a stable, negative feedback that works to equalize the growth rate everywhere, and vertical orientation is maintained [4]. In many cases Ni nanoparticles were found at the top of carbon nanotubes which may indicate “tip-growth” mechanism of carbon nanotubes [2–4]. Additionally, we observe Ni nanoparticles with elongated shape placed inside the carbon nanotubes. Both kinds are visible in the TEM images as dark objects (Fig. 6a) and in SEM images as white objects in LABE mode (Fig. 5b). Additionally Ni nanoparticles seen on top of CNT are covered with multiple planes of graphite (Fig. 6c, 6d).

Mirosław Kozłowski, Piotr Dłużewski, Ewa Kowalska, Elżbieta Czerwosz

Figure 5.

MWCNTs with Ni nanoparticles a) SE mode, b) LABE mode.

Figure 6.

a, c) TEM image of MWCNT, b) diffraction pattern, d) {0002} planes of graphite coating of the Ni nanoparticles.

The number of walls observed is higher than 10, which could be connected with the tip-growth mechanism [6]. Ni nanoparticles placed inside MWCNT could suggest that

in case of our two-steps process (which consists of a fast 10 min. CVD process), a coexisting mechanism of growth (base-growth) is also active. The Ni nanoparticle posi-

347

Secondary electron microscopy and transmission electron microscopy studies of carbon nanotubes in C-Ni films

tions inside CNT suggest that this particle could be only a fragment of a larger particle which is located at the base of the CNT. The short run time of the CVD process and an average MWCNT growth allow the MWCNT growth time to be determined as ∼ 0.5 nm/s. In Fig. 6d the distance between planes of graphite coating of Ni nanoparticles is shown as 0.33 nm which remains in good agreement with [11]. Such embedding protects Ni nanoparticles from oxidation in the atmosphere and explains the difference in diameter between top Ni nanoparticles determined in SE mode and LABE mode in SEM investigations (Fig. 5). SAED patterns are shown in Fig. 6a. These patterns present concentric of weak intensity rings from small Ni nanoparticles and rings with strong intensity from graphite. In Fig. 6b the origin of rings is marked with yellow and blue colors respectively.

6.

Conclusion

MWCNTs obtained by the two step method with a Ni catalyst were characterized by SEM and TEM. These investigations allowed the topography and morphology of MWCNTs films th be determined, as well as the distribution of these MWCNTs on the film surface. In this two step method we can obtain carbonaceous film in which short, curled and isolated MWCNTs are present.

348

Acknowledgements These investigations were supported from project financed by Polish Ministry of Science and Education (No. ERANET 400/ERA-NET/2009).

References [1] Th.D. Makris, L. Giorgi, R. Giorgi, N. Lisi, E. Salernitano, Diamond Relat. Mater. 14, 815 (2005) [2] M. Meyyappan, L. Delzeit, A. Cassel, D. Hash, Plasma Sources Sci. Technol. 12, 205 (2003) [3] V. Merkulov, A.V. Melechko, M.A. Guillorn, D.H. Lowndes, M.L. Simpson, Appl. Phys. Lett. 79, 2970 (2001) [4] J. Han et al., J. Appl. Phys. 91, 483 (2002) [5] B-R. Huang, T-C. Lin, Y-K. Yang, S-D. Tzeng, Diamond Relat. Mater. 19, 158 (2010) [6] S. Chang, T. ChaiLin, T. SingLi, S. Huang, Microelectron. J. 39, 1572 (2008) [7] C. Bittencourt et al., Chem. Phys. Lett. 436, 368 (2007) [8] L. Li, K. Lafdi, Sens. Actuators B 132, 202 (2008) [9] A. Gohiera, C.P. Ewels, T.M. Minea, M.A. Djouadi, Carbon 46, 1331 (2008) [10] E. Czerwosz, E. Kowalska, H. Wronka, J. Radomska, patent notification 2008 nr P384 591 [11] ICDD No. 04-007-8496, graphite, syn